Developing a scalable hybrid MPI/OpenMP unstructured finite element model

•Threaded preconditioner have been experimented.•Investigated the multigrid performance in detail.•The code can do the strong scaling up to 16 thousands cores.•Reviewed achievements and lessons for overall OpenMP implementation. The trend of all modern computer architectures, and the path to exascal...

Full description

Saved in:

Bibliographic Details
Published in:	Computers & fluids Vol. 110; pp. 227 - 234
Main Authors:	Guo, Xiaohu, Lange, Michael, Gorman, Gerard, Mitchell, Lawrence, Weiland, Michèle
Format:	Journal Article
Language:	English
Published:	Elsevier Ltd 01.03.2015
Subjects:	Algorithms Evolutionary FEM Finite element method Fluidity-ICOM HYPRE Mathematical analysis Mathematical models Matrix assembly Message passing MPI OpenMP PETSc Platforms Sparse linear solver SpMV Threaded SpMV HYPRE OpenMP MPI Sparse linear solver PETSc Fluidity-ICOM Matrix assembly FEM
ISSN:	0045-7930, 1879-0747
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!